Index of Agreement

Problem

Compute Willmott's index of agreement: 1 - sum((forecast - actual)^2) / sum((abs(forecast - mean_actual) + abs(actual - mean_actual))^2).

Return 0.0 when the denominator is zero or actual is empty.

Hint

The index of agreement is bounded and less sensitive to extreme errors than R squared.

case 1

in: [[1,2,3,4],[1,2,3,4]]

expected: 1

case 2

in: [[1,2,3,4],[2,2,2,2]]

expected: 0.4

case 3

in: [[1,2],[2,1]]

expected: 0

case 4

in: [[5,5,5],[5,5,5]]

expected: 0

Starter codePython

def index_of_agreement(actual, forecast):
    # Your code here
    pass

4 test cases · run your code against them in the editor